US9395920B2 - Throttle disk I/O using disk drive simulation model - Google Patents

Throttle disk I/O using disk drive simulation model Download PDF

Info

Publication number
US9395920B2
US9395920B2 US13/298,993 US201113298993A US9395920B2 US 9395920 B2 US9395920 B2 US 9395920B2 US 201113298993 A US201113298993 A US 201113298993A US 9395920 B2 US9395920 B2 US 9395920B2
Authority
US
United States
Prior art keywords
disk
request
hard
execution time
drive
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active, expires
Application number
US13/298,993
Other languages
English (en)
Other versions
US20130132057A1 (en
Inventor
Yimin Deng
Ho Yuen Chau
Yue Zuo
Forrest Curtis Foltz
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Microsoft Technology Licensing LLC
Original Assignee
Microsoft Technology Licensing LLC
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Microsoft Technology Licensing LLC filed Critical Microsoft Technology Licensing LLC
Assigned to MICROSOFT CORPORATION reassignment MICROSOFT CORPORATION ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: CHAU, HO YUEN, DENG, YIMIN, FOLTZ, FORREST CURTIS, ZUO, Yue
Priority to US13/298,993 priority Critical patent/US9395920B2/en
Priority to PCT/US2012/064587 priority patent/WO2013074426A1/en
Priority to ES12850595.5T priority patent/ES2628611T3/es
Priority to EP12850595.5A priority patent/EP2780799B1/en
Priority to CN201210465126.7A priority patent/CN103116473B/zh
Publication of US20130132057A1 publication Critical patent/US20130132057A1/en
Assigned to MICROSOFT TECHNOLOGY LICENSING, LLC reassignment MICROSOFT TECHNOLOGY LICENSING, LLC ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: MICROSOFT CORPORATION
Publication of US9395920B2 publication Critical patent/US9395920B2/en
Application granted granted Critical
Active legal-status Critical Current
Adjusted expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
    • G06F3/061—Improving I/O performance
    • G06F3/0613—Improving I/O performance in relation to throughput
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
    • G06F11/3409—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment for performance assessment
    • G06F11/3433—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment for performance assessment for load management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
    • G06F11/3466—Performance evaluation by tracing or monitoring
    • G06F11/3485—Performance evaluation by tracing or monitoring for I/O devices
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
    • G06F3/0655—Vertical data movement, i.e. input-output transfer; data movement between one or more hosts and one or more storage devices
    • G06F3/0659—Command handling arrangements, e.g. command buffers, queues, command scheduling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
    • G06F3/0671—In-line storage system
    • G06F3/0673—Single storage device
    • G06F3/0674—Disk device
    • G06F3/0676—Magnetic disk device
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
    • G06F11/3409—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment for performance assessment
    • G06F11/3419—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment for performance assessment by assessing time
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
    • G06F11/3447—Performance evaluation by modeling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2201/00—Indexing scheme relating to error detection, to error correction, and to monitoring
    • G06F2201/815—Virtual
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
    • G06F3/0662—Virtualisation aspects
    • G06F3/0664—Virtualisation aspects at device level, e.g. emulation of a storage device or system

Definitions

  • a data center e.g., physical cloud computing infrastructure
  • services e.g., web applications, email services, search engine services, etc.
  • These large-scale networked systems typically include a large number of resources distributed throughout the data center, in which each resource resembles a physical machine or a virtual machine (VM) running on a physical node or host.
  • VM virtual machine
  • the data center hosts multiple tenants (e.g., customer programs), these resources are optimally allocated from the same data center to the different tenants.
  • VMs will concurrently run on the same physical node within a computing network, or the data center.
  • These VMs that share a common physical node may be allocated to the different tenants and may require different amounts of resources at various times.
  • the resources may include a physical disk (e.g., hard-disk driver) associated with a physical node, where the physical disk has a limited amount of accessibility over a certain time frame.
  • the tenant's VMs running on the physical node can potentially stress the physical disk by sending a large amount of requests thereto, preventing other VMs running on the physical node from fairly sharing the resources thereof.
  • disk input/output (I/O) performance e.g., time to access a segment of data
  • CPU-performance e.g., processor speed
  • network performance e.g., bandwidth usage
  • this approach considers a number of times a disk head is asked to move over the physical disk within a given time frame. That is, using the requests/sec approach does not consider the size of data that is being accessed upon movement of the disk head.
  • the MB/sec-based approach relevant to sequential access only
  • the request/sec-based approach relevant to random access only
  • these existing approaches fail to account for both sequential access and random access on the physical disk.
  • the MB/sec-based approach is focused on sequential accesses, thus, assuming the risk of being overwhelmed by a flood of random accesses and generating a backlog of requests.
  • the request/sec-based approach is focused on random accesses, thus, assuming the risk of encountering a string of sequential accesses and unnecessarily limiting the amount of throughput to the physical disk. Consequently, failing to account for sequential access and random access, or to recognize the dramatic differences in resultant disk I/O performance between them, renders these approaches ineffective.
  • Embodiments of the present invention provide a mechanism that throttles requests issued from service application(s) running on virtual machines (VMs), which reside on a common physical machine or node of a data center, in accordance with a disk-drive model.
  • the phrase “disk-drive model” generally refers to a soft-disk representation of a hard-disk drive (HDD), where the disk-drive model simulates performance characteristics of the HDD when computing projected execution times for carrying out the service-application requests.
  • the disk-drive model simulates performance characteristics of the HDD by considering at least the following two factors of the HDD: a set of parameters that correspond to a set of actual parameters presently existing on the HDD; and a current state of the disk head of the HDD.
  • the current state of the disk head can be used to calculate a time consumed by moving the disk head from a first position that satisfies a presently executing request to a second position that satisfies a pending request.
  • Previous approaches for quantifying performance of an HDD are limited to considering either MB/sec (amount of information transferred or request/sec (rate of operation).
  • the approach employed in embodiments of the present invention relies on the disk-drive model that accurately quantifies disk I/O performance based on various factors, including the two factors mentioned above. Further, the approach in embodiments of the present invention applies the disk-drive model to determine how the disk I/O requests to read or write (hereinafter “requests”) should be handled, such as limiting, or “throttling,” the requests based on the fact of how the HDD is performing relative to the disk-drive model.
  • Properly throttling disk I/O is advantageous in a cloud-computing environment for various reasons, such as providing a way to decouple VM Disk performance from underlying physical storage hardware (e.g., HDD) and providing the ability to assign a performance threshold dynamically to individual VMs.
  • the performance threshold is calculated upon comparing a projected execution time of a request (using the disk-drive model) and an actual execution time of the request (measuring the time taken by the HDD to carry out the request).
  • the disk-drive model is used for the purpose of throttling the requests for disk I/O. Initially, for each disk I/O operation being request, a projected execution time (i.e., time spent by the disk-drive model to handle the disk I/O operation) is computed. In operation, the projected execution time spent by the disk-drive model may be used as a baseline bar for the throttling decision.
  • a projected execution time i.e., time spent by the disk-drive model to handle the disk I/O operation
  • an actual execution time i.e., time spent by the HDD to carry out the disk I/O operation
  • the HDD is outperformed.
  • the subsequent disk I/O operations may be executed in a delayed manner, or, in other words, the subsequent requests may be throttled.
  • throttling may involve adding time when scheduling the subsequent requests when the HDD is outperformed.
  • throttling may involve processing the subsequent requests immediately when the HDD is underperformed.
  • throttling may be used when the HDD is not over stressed. In one example, if different performance levels are assigned to different types of VMs, throttling may help enforce these performance levels are upheld. In another example, if a particular customer's VM is guaranteed a performance level regardless of where it is deployed (e.g., on a fast physical machine or on a slower machine), throttling may help ensure that guaranteed performance level is upheld.
  • FIG. 1 is a block diagram of an exemplary computing environment suitable for use in implementing embodiments of the present invention
  • FIG. 2 is a block diagram illustrating an exemplary cloud-computing platform, suitable for use in implementing embodiments of the present invention, that is configured to employ a disk-drive model for computing a projected execution time of an incoming request;
  • FIG. 3 is block diagram of an exemplary distributed computing environment depicting queues within a physical node communicating to a filter driver, in accordance with an embodiment of the present invention
  • FIG. 4 is a flow diagram showing an exemplary method for throttling requests from one or more virtual machines (VMs) to a hard-disk drive (HDD) using the filter driver, in accordance with an embodiment of the present invention.
  • VMs virtual machines
  • HDD hard-disk drive
  • FIG. 5 is a flow diagram showing an exemplary method for limiting a rate of requests flowing to the HDD, in accordance with an embodiment of the present invention.
  • Embodiments of the present invention relate to methods, computer systems, and computer-readable media for applying a throttle to a source (e.g., virtual machines (VMs) instantiated on a physical node) of requests for disk I/O operations to be handled by a hard-disk drive (HDD).
  • the applied throttle may, in operation, meter the requests from the VMs according a comparison between a projected execution time of a request (using the disk-drive model) and an actual execution time of the request (measuring the time taken by the HDD to carry out the request).
  • metering involves delaying delivery of those requests that are issues after identifying the projected execution time exceeds the actual execution time.
  • metering may be implemented by a filter driver that causes the host to store, or “enqueue,” the delayed requests issued by the VMs.
  • embodiments of the present invention relate to one or more computer-readable media having computer-executable instructions embodied thereon that, when executed, perform a method for throttling requests from one or more VMs to the HDD using the filter driver.
  • the method may commence upon the filter driver receiving a subject request from a service application running on one or more of the VMs.
  • the disk-drive model is accessed.
  • the disk-drive model behaves substantially similar to the HDD.
  • the process of accessing includes reading the disk-drive model's simulation of performance characteristics of the hard-disk drive and reading the disk-drive model's simulation of a current state of a disk head of the hard-disk drive.
  • the method carried out by the filter driver may further include the step of computing a projected execution time to carry out the subject request as a function of the performance characteristics and the current state of the disk head. Also, the actual execution time to carry out the subject request may be measured upon the filter driver allowing the subject request to pass to the HDD. Using a comparison of the projected execution time and the actual execution time, the traffic of the requests from the VMs, which are issued after the subject request to the HDD, are throttled.
  • embodiments of the present invention relate to a computer system capable of throttling a rate at which requests from a service application are executed by the HDD, where throttling is managed in accordance with the disk-drive model.
  • the computer system includes at least one physical machine (e.g., node of a data center) configured with a computer-storage medium that has a plurality of computer software components running thereon.
  • the computer software components include a VM, a VM bus, and a file system driver running in the host partition.
  • the VM is configured for hosting a portion of the service application (e.g., an instance of a service-application role or component program).
  • This portion of the service application is capable of issuing requests to read data from or write data to a virtual hard-disk (VHD) file associated with the VM.
  • VHD virtual hard-disk
  • the filter driver is configured for comparing a projected execution time of a first request against an actual execution time of the first request.
  • the “projected execution time” is calculated using the disk-drive model that simulates performance characteristics of the hard-disk drive, while the “actual execution time” is measured upon completion of a request at the hard-disk drive.
  • the filter driver is further configured for affecting a timing for carrying out the subsequent request based on an outcome of the comparison.
  • the VM bus is configured to deliver the request from the service application to the filter driver and the queue(s) are configured to hold requests issued from the VM that were delayed by the filter driver, Typically, the requests in these queues will be processed and sent to the HDD at a later time.
  • embodiments of the present invention relate to a computerized method for limiting a rate of requests flowing to a hard-disk drive.
  • the method involves employing the disk-drive model to compute the projected execution time for the HDD to carry out a first request and measuring the actual execution time taken by the HDD to carry out the first request.
  • the disk-drive model considers a time for a disk head (used to read data from or write data to a hard disk of the HDD) to travel from the last sector of data targeted by the first request and to the first sector of data targeted by a second request.
  • the projected execution time is compared against the actual execution time and the rate of the requests allowed to pass to the HDD is managed as a function of an outcome of the comparison.
  • the process of managing includes imposing a delay upon the request before passing the request to the HDD when the projected execution time exceeds the actual execution time by a value greater than a performance threshold.
  • the process of managing includes passing the request to the HDD upon receipt of the request when projected execution time does not exceed the actual execution time by a value greater than the performance threshold.
  • an exemplary operating environment suitable for implementing embodiments of the present invention is described below.
  • an exemplary operating environment supports functionality of a mechanism (e.g., filter driver 230 of FIG. 2 ) that allows VMs residing on a common physical node to each send requests for disk I/O by placing restrictions on resource consumption at the HDD; thus, ameliorating a stressed condition of overloading the HDD with pending requests or adversely affecting throughput to the HDD unnecessarily.
  • a mechanism e.g., filter driver 230 of FIG. 2
  • computing device 100 an exemplary operating environment for implementing embodiments of the present invention is shown and designated generally as computing device 100 .
  • Computing device 100 is but one example of a suitable computing environment and is not intended to suggest any limitation as to the scope of use or functionality of the invention. Neither should the computing device 100 be interpreted as having any dependency or requirement relating to any one or combination of components illustrated.
  • the invention may be described in the general context of computer code or machine-useable instructions, including computer-executable instructions such as program modules, being executed by a computer or other machine, such as a personal data assistant or other handheld device.
  • program modules including routines, programs, objects, components, data structures, etc., refer to code that performs particular tasks or implements particular abstract data types.
  • the invention may be practiced in a variety of system configurations, including handheld devices, consumer electronics, general-purpose computers, more specialty computing devices, etc.
  • the invention may also be practiced in distributed computing environments where tasks are performed by remote-processing devices that are linked through a communications network.
  • computing device 100 includes a bus 110 that directly or indirectly couples the following devices: memory 112 , one or more processors 114 , one or more presentation components 116 , input/output (I/O) ports 118 , input/output components 120 , and an illustrative power supply 122 .
  • Bus 110 represents what may be one or more busses (such as an address bus, data bus, or combination thereof).
  • FIG. 1 is merely illustrative of an exemplary computing device that can be used in connection with one or more embodiments of the present invention. Distinction is not made between such categories as “workstation,” “server,” “laptop,” “handheld device,” etc., as all are contemplated within the scope of FIG. 1 and reference to “computing device.”
  • Computer-readable media can be any available media that can be accessed by computing device 100 and includes both volatile and nonvolatile media, removable and nonremovable media.
  • Computer-readable media may comprise computer storage media and communication media.
  • Computer storage media include both volatile and nonvolatile, removable and nonremovable media implemented in any method or technology for storage of information such as computer-readable instructions, data structures, program modules, or other data.
  • Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium, which can be used to store the desired information and which can be accessed by computing device 100 .
  • Communication media typically embody computer-readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and include any information delivery media.
  • modulated data signal means a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal.
  • communication media include wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared, and other wireless media. Combinations of any of the above should also be included within the scope of computer-readable media.
  • Memory 112 includes computer-storage media in the form of volatile and/or nonvolatile memory.
  • the memory may be removable, nonremovable, or a combination thereof.
  • Exemplary hardware devices include solid-state memory, hard drives, optical-disc drives, etc.
  • Computing device 100 includes one or more processors that read data from various entities such as memory 112 or I/O components 120 .
  • Presentation component(s) 116 present data indications to a user or other device.
  • Exemplary presentation components include a display device, speaker, printing component, vibrating component, etc.
  • I/O ports 118 allow computing device 100 to be logically coupled to other devices including I/O components 120 , some of which may be built in.
  • I/O components 120 include a microphone, joystick, game pad, satellite dish, scanner, printer, wireless device, etc.
  • a physical node 215 may be implemented by the exemplary computing device 100 of FIG. 1 .
  • One or more of the virtual machines (VMs) 201 , 202 , 203 , and 204 may include portions of the memory 112 of FIG. 1 and/or portions of the processors 114 of FIG. 1 .
  • the VMs 201 - 204 may each represent a respective guest partition of the node 215 , where the guest partition is enabled to host a service application, or, at least, a portion thereof.
  • the service application is owned by a tenant (e.g., customer) of a cloud-computing platform that hosts the service application in a distributed manner.
  • the node 215 may further support a host 210 with various component 220 , 225 , 230 , 235 , 240 , and 245 running thereon.
  • the host 210 represents a root partition of the node 215 where the root partition is tasked with generally managing requests from the guest partitions, such as requests to implement disk I/O operations at the hard-disk drive (HDD) 260 .
  • HDD hard-disk drive
  • FIG. 2 a block diagram is illustrated, in accordance with an embodiment of the present invention, showing an exemplary cloud-computing platform 200 that is configured to employ a disk-drive model 235 for computing a projected execution time of an incoming request 270 .
  • the request 270 is issued from a role instance 211 of a service application that is assigned to the first VM 201 .
  • a VM bus 205 is employed to convey the request 270 from the first VM 201 to the host 210 .
  • the VM bus 205 is used to transport traffic of requests from the VMs 201 - 204 on the node 215 to the host 210 for throttling prior to processing at the HDD 260 .
  • FIG. 2 Although only one service-application portion (role instance 211 ) is illustrated in FIG. 2 , it should be understood that various other component programs of other service applications may be assigned to and placed upon the VMs 201 - 204 , as discussed more fully below. That is, the VMs 201 - 204 within the node 215 may be occupied by various service applications.
  • a portion of service application A may occupy VM 202 while portions of service application B may occupy VMs 203 and 204 , where each of the VMs 202 - 204 that are instantiated on the physical node 215 to support functionality of the service applications A and B.
  • the cloud-computing platform 200 shown in FIG. 2 is merely an example of one suitable computing system environment and is not intended to suggest any limitation as to the scope of use or functionality of embodiments of the present invention.
  • the cloud-computing platform 200 may be a public cloud, a private cloud, or a dedicated cloud. Neither should the cloud-computing platform 200 be interpreted as having any dependency or requirement related to any single component or combination of components illustrated therein.
  • the various blocks of FIG. 2 are shown with lines for the sake of clarity, in reality, delineating various components is not so clear, and metaphorically, the lines would more accurately be grey and fuzzy.
  • any number of physical machines, virtual machines, data centers, endpoints, or combinations thereof may be employed to achieve the desired functionality within the scope of embodiments of the present invention.
  • the cloud-computing platform 200 typically includes data centers configured to host and support operation of the VMs 201 - 204 that host portions (e.g., role instance 211 or other endpoints) of distributed service applications owned by tenants/customers of the cloud-computing platform 200 .
  • the role instance 211 may represent a portion of software or a component program that participates in supporting functionality of the service application. It will be understood and appreciated that role instance 211 shown in FIG. 2 is merely an example of a suitable part to support a service application and is not intended to suggest any limitation as to the scope of use or functionality of embodiments of the present invention.
  • VMs 201 - 204 are instantiated and designated to one or more service applications, respectively, based on demands (e.g., amount of processing load) placed on the service application.
  • demands e.g., amount of processing load
  • the phrase “virtual machine” is not meant to be limiting, and may refer to any software, application, operating system, or program that is executed by a processing unit to underlie the functionality of the service application.
  • the VMs 201 - 204 may request processing capacity, access to storage locations, resources, and other assets (e.g., hosting agent) within the node 215 to properly support the service applications.
  • the VMs 201 - 204 may be dynamically allocated resources (e.g., network bandwidth, CPU processing-capacity, or disk I/O at the HDD) to satisfy a current workload.
  • the host 210 may assist with allocating resources to the VMs 201 - 204 .
  • the filter driver 230 of the host 210 may manage throttling of requests issued from the VMs 201 - 204 that are directed to the HDD 260 for processing.
  • the VMs 201 - 204 may be dynamically allocated resources within the node 215 .
  • the node 215 may represent any form of computing device, such as, for example, a personal computer, a desktop computer, a laptop computer, a mobile device, a consumer electronic device, server(s), the computing device 100 of FIG. 1 , and the like.
  • the node 215 supports the operations of the VMs 201 - 204 , respectively, while simultaneously hosting other virtual machines (not shown) carved out for supporting other tenants of cloud-computing platform 200 , where the tenants include endpoints of other service applications owned by different customers.
  • the node 215 includes, or is linked to, some form of a computing unit (e.g., central processing unit, microprocessor, etc.) to support operations of the VMs 201 - 204 , and/or components 220 , 225 , 230 , 235 , 240 , and 245 running thereon.
  • a computing unit e.g., central processing unit, microprocessor, etc.
  • the phrase “computing unit” generally refers to a dedicated computing device with processing power and storage memory, which supports one or more operating systems or other underlying software.
  • the computing unit is configured with tangible hardware elements, or machines, that are integral, or operably coupled, to the node 215 to enable each endpoint to perform a variety of processes and operations.
  • the computing unit may encompass a processor (not shown) coupled to the computer-readable medium accommodated by the node 215 .
  • the computer-readable medium stores, at least temporarily, a plurality of computer software components (e.g., components 220 , 225 , 230 , 235 , 240 , and 245 ) that are executable by the processor.
  • the term “processor” is not meant to be limiting and may encompass any elements of the computing unit that act in a computational capacity. In such capacity, the processor may be configured as a tangible article that processes instructions. In an exemplary embodiment, processing may involve fetching, decoding/interpreting, executing, and writing back instructions.
  • the VMs 201 - 204 operate within the context of the cloud-computing platform 200 and, accordingly, may communicate internally within the node 215 , communicate across physical nodes through connections dynamically made that traverse a data center, and externally through a physical network topology to resources of a remote network (e.g., enterprise private network).
  • the connections may involve interconnecting VMs distributed across physical resources of the data center via a network cloud (not shown).
  • the network cloud interconnects these resources such that one VM may recognize a location of another VM in order to establish a communication therebetween. For instance, the network cloud may establish this communication over channels connecting VMs hosting component programs of a common service application.
  • the channels may include, without limitation, one or more local area networks (LANs) and/or wide area networks (WANs).
  • LANs local area networks
  • WANs wide area networks
  • Such networking environments are commonplace in offices, enterprise-wide computer networks, intranets, and the Internet. Accordingly, the network is not further described herein.
  • the I/O manager 220 initially receives the request 270 and parses the request to gain an understanding of the instructions therein.
  • the instructions of the request 270 may include a command to carry out a read operation or a write operation that consumes disk I/O at the HDD 260 .
  • the I/O manager 220 then passes the request 270 to the filter manager 225 for determining whether to handle the request 270 immediately or to impose a delay on the request 270 .
  • the filter manager 225 may have a plurality of types of filters at its disposal to assist in properly managing the traffic of requests from the VMs 201 - 204 .
  • the filter manager may consult the filter driver 230 to throttle the traffic of requests from the VMs 201 - 204 , in accordance with a disk-drive model 235 and a current status of the HDD 260 . Embodiments of throttling will be discussed more fully below with respect to FIG. 3 .
  • the file system 240 interprets the request 270 , determines the appropriate virtual hard disk (VHD) file to access, and identifies a location of the VHD file.
  • the storage driver stack 245 implements the request 270 by controlling the HDD 260 to read from or write to the VHD file on a hard disk of the HDD 260 .
  • VHD virtual hard disk
  • the VHD file generally represents a virtual hard-disk file with a particular formatting applied thereto. This file formatting can maintain proper storage of data on the HDD 260 , such as disk partitions and a file system. These disk partitions may include structured data, such as files and folders. In operation, the VHD file may be used as the hard-disk memory of a virtual machine.
  • the HDD 260 represents a nonvolatile, random-access digital magnetic storage device. Included within the HDD 260 , is a rotating rigid platter (e.g., a hard disk) that is rotated by a motor-driven spindle within a protective enclosure. Data is magnetically read from and written to the hard disk by at least one disk head.
  • the HDD 260 employs a mechanism that moves the disk head to a correct location and reads a number of bytes from the hard disk. Moving the head to the correct location consumes a certain amount of time that is considered by the filter driver 230 when assigning tasks to HDD 260 .
  • the request 270 will ask the HDD 260 to target a particular portion of data on the hard disk as part of a read operation or a write operation.
  • the targeted data may be located within a first sector of the hard disk while another portion of data targeted by a subsequent request may be located within the second sector of the hard disk.
  • the first sector and the second sector may be located on different tracks (e.g., concentric circles of magnetically stored information) recorded on the hard disk. Accordingly, upon completing the request 270 and commencing carrying out the subsequent request, a certain amount of time is consumed when the disk head travels a lateral distance from the first sector to the second sector of the hard disk.
  • the sectors holding data on the magnetic surface of the hard disk represent micrometer-sized subdivisions or magnetic domains that are assigned to a particular file or folder (e.g., VHD file).
  • the HDD 260 undertakes several processes that consume time and, thus, affect scheduling of the requests from the VMs 201 - 204 .
  • These time-consuming processes of the HDD 260 include controller latency, seek speed, rotational latency, and data-transfer rate.
  • the controller latency refers to the overhead that the disk controller (e.g., I/O manager 220 ) imposes in interpreting and executing the request 270 .
  • the seek speed refers to the time consumed to move the disk head from a track on the hard disk holding the first sector of data (targeted by a pending request) to another track on the disk head holding the second sector of data (targeted by a subsequent request).
  • the rotational latency refers to the time consumed to rotate the hard disk to a beginning of the second sector of data once the disk head is moved to the appropriate track.
  • the data-transfer rate refers to the speed at which the hard disk is rotated and at which the data is read from or written to the hard disk (e.g., number of blocks of contiguous files transferred to or from the hard disk).
  • these time-consuming processes are quantified as a set of actual parameters affecting disk I/O performance of the HDD 260 . That is, this set of actual parameters exposes performance characteristics of the HDD 260 that can be employed to influence the throttling of requests directed to the HDD 260 .
  • a set of estimated parameters substantially corresponding to the set of actual parameters of the HDD 260 ) that attempt to specify the performance characteristics of the HDD 260 may be incorporated within the disk-drive model 235 .
  • the phrase “disk-drive model” is not meant to be limiting but may encompass any logic, heuristics, or software-based rules that assist the filter driver 230 in managing the flow of the traffic directed toward the HDD 260 .
  • the disk-drive model 235 is built by the filter driver 230 to substantially mimic behavior of the HDD 260 by programming the set of estimated parameters (discussed above) within the disk-drive model 235 . Further, the disk-drive model 235 may be programmed to consider a current state of the HDD 260 , which specifies the present disk-head position and total time consumed in satisfying the pending request 270 .
  • the current state of the HDD 260 may be used to determine the time required for the disk head to travel to the next location on the hard disk targeted by the subsequent request (e.g., time consumed to cover a distance of travel when updating the position of the disk head).
  • the time consumed is estimated based on the type of access (read or write), the location of the data, and amount of processing required to complete the request 270 .
  • the projected execution time determined using the disk-drive model 235 used by the filter driver 230 to govern the throttling of requests, considers both the set of actual parameters (capturing present performance characteristics of the HDD 260 ) as well as the time consumed to transition from one request to another (using the calculation of the distance between a current disk-head position and a next disk-head position).
  • sequential access involves reading the sectors of the hard disk consecutively within a particular track. That is, sequential access comprises reading one adjacent sector after another physically proximate sector on the hard disk without translating the disk head. Typically, there is a minimal amount of time consumed by moving the disk head to the next physically proximate sector on the hard disk.
  • Random access involves reading data in different sectors that are in various locations (e.g., different tracks on the hard disk) in relation to each other. Typically, there is a significant amount of time consumed when the disk head travels from one sector to another.
  • the disk-drive model's 235 consideration of the set of parameters estimated from the HDD 260 and the current/future state of the disk-head position accounts for time consumption involved in both sequential access and random access, respectively.
  • FIG. 3 a block diagram is illustrated showing an exemplary distributed computing environment 300 depicting queues 301 - 304 within one or more physical node(s) communicating with the filter driver 230 , in accordance with an embodiment of the present invention.
  • the VMs 201 - 204 manage the queues 301 - 304 , respectively, although each of the VMs 201 - 204 may have multiple queues for temporarily holding pending requests. Once released from the queues 301 - 304 , the requests are serviced by the file system in coordination with the HDD 260 . It should be noted that each of the queues 301 - 304 may be managed separately, where each is assigned a percentage of the software disk.
  • service applications running within the VMs 201 - 204 may issue requests to perform a read and/or write operation on the HDD 260 .
  • these read and/or write operations may be directed to a host over a VM bus and may be targeted to respective VHD files of interest that reside on the HDD 260 .
  • the filter driver 230 intercepts each read and write operation and processes them before being passed to the file system.
  • the filter driver 230 is responsible for the management of throttling disk I/O requests.
  • the filter driver 230 computes the time required for the disk-drive model to process the read or write operation that satisfies the request from the service application. This computation considers at least the following two inputs: a current state of the disk-drive model; and attributes of the currently executed read or write operation (e.g., data size and its location).
  • the time expended to process the read or write operation by the disk-drive model may be computed as the sum of the latencies listed below: controller latency (e.g., time estimated for the disk controller to parse the request); seek time (e.g., time estimated for the disk head to move from a current position on a first HDD track to a projected position on a second HDD track, where data to be accessed is located); rotational latency (e.g., time estimated for the requested sector to rotate into a position under the disk head; and data transfer time (e.g., time estimated to read or write the data of a given size).
  • controller latency e.g., time estimated for the disk controller to parse the request
  • seek time e.g., time estimated for the disk head to move from a current position on a first HDD track to a projected position on a second HDD track, where data to be accessed is located
  • rotational latency e.g., time estimated for the requested sector to rotate into a position under
  • the filter driver 230 may record the actual time consumed by the physical disk hardware to process a subject request.
  • the filter driver 230 may implement an analysis that determines the difference in time between the two metrics above (projected execution time and actual execution time). This difference in time indicates whether the physical disk hardware is outperforming or underperforming the disk-drive model. If the physical disk hardware outperforms the disk-drive model then the request may be delayed by holding it in a queue associated with the VHD file of the requesting VM requesting VM. Typically, there is a separate queue to hold the delayed requests for each VHD. A VM may be assigned more than one VHD, thus, the VM may be associated with more than one queue.
  • the filter driver 230 may elect to delay servicing requests issued from VM 201 .
  • the amount of delay of the request may be based on the difference of the comparison (e.g., if the physical disk hardware performs much faster, then a longer delay is imposed). Accordingly, the filter driver attempts to synch the times predicted by the disk-drive model and the time actually taken by the HDD to process the request.
  • the filter driver 230 lifts the delays on the requests.
  • the filter driver 230 can both speed up or can slow down the flow of requests to mimic the performance characteristics of the HDD.
  • the filter driver 230 is equipped to adapt to a level of performance of the HDD 260 , whether the HDD 260 represents physical disk hardware on a personal desktop or whether the HDD 260 represents a commercial-grade disk employed in a data center.
  • the filter driver 230 may perform various callback routines to implement these duties.
  • One of the callback routines involves a preoperation callback that is invoked by the filter manager 230 for each disk I/O request before the request is processed by the HDD 260 .
  • the preoperation callback initially, determines whether the request should be handled immediately. This decision is based on the difference of the total time consumed by the physical disk hardware and the total projected time to handle the same requests by using the disk-drive model. If the decision indicates the request should be handled immediately, the filter driver 230 passes the request through to the file system and storage driver stack (see reference numerals 240 and 245 of FIG. 2 ).
  • the filter driver 230 inserts the request into an appropriate queue and schedules a timer routine that triggers processing of the queue at a later time.
  • a timer routine that triggers processing of the queue at a later time.
  • Another of the callback routines involves a postoperation callback that is invoked by the filter manager 230 for each disk I/O request after the request is processed.
  • the post-operation callback records the time consumed by the physical disk hardware to handle the request and updates the total time consumed by the physical disk hardware. Further, the postoperation callback computes the projected execution time to handle the request by accessing the disk-drive model. The projected execution time is used to update the total time spent by the disk-drive model.
  • the current state of the HDD 260 such as disk-head position, may be stored for future reference in the disk-drive model.
  • FIG. 4 a flow diagram is illustrated that shows an exemplary method 400 for throttling requests from one or more VMs to the HDD using the filter driver, in accordance with an embodiment of the present invention.
  • the terms “step” and/or “block” may be used herein to connote different elements of methods employed, the terms should not be interpreted as implying any particular order among or between various steps herein disclosed unless and except when the order of individual steps is explicitly described.
  • the method 400 may commence upon the filter driver receiving a subject request from a service application running on the VMs, as indicated at block 410 .
  • the subject request may represent a read operation or a write operation directed to a VHD file that maintains stored data on a hard disk of the HDD, wherein the VHD file is associated with at least one of the VMs.
  • a disk head of the HDD is configured to read or write to the hard disk in response to the read operation or the write operation, respectively.
  • the disk-drive model Upon parsing the request at the filter driver, the disk-drive model is accessed, as indicated at block 412 .
  • the disk-drive model behaves substantially similar to the HDD and embodies a soft-disk representation of the HDD.
  • the process of accessing includes reading the disk-drive model's simulation of performance characteristics of the hard-disk drive (see block 414 ) and reading the disk-drive model's simulation of a current state of a disk head of the hard-disk drive (see block 416 ).
  • the simulated performance characteristics are specified by a set of estimated parameters that correspond to a set of actual parameters existing on the hard-disk drive.
  • the set of estimated parameters includes at least the following: a latency of a controller; a speed at which the disk head seeks; a latency of rotation; and a rate of data transfer.
  • the controller latency includes the latency that the disk controller imposes in interpreting and executing a request command.
  • the seek speed includes a rate at which the disk head travels across tracks on a hard disk.
  • the rotational latency includes a time to rotate a hard disk to a location of data targeted by the subject request.
  • the data transfer rate includes a speed at which the disk head reads data targeted by the subject request from a hard disk.
  • the method 400 carried out by the filter driver may further include the step of computing a projected execution time to carry out the subject request as a function of the performance characteristics and the current state of the disk head, as indicated at block 418 .
  • computing a projected execution time to carry out the subject request involves at least the following steps: identifying a first position of the disk head when carrying out a precedent request; extrapolating from the disk-drive model a second position of the disk head when carrying out the subject request; and calculating a distance of travel between the first position and the second position.
  • the actual execution time to carry out the subject request may be measured upon the filter driver allowing the subject request to pass to the HDD, as indicated at block 420 .
  • the traffic of the requests from the VMs, which are issued after the subject request to the HDD are throttled, as indicated at block 422 .
  • throttling traffic of the requests from the VMs to the HDD using a comparison of the projected execution time and the actual execution time involves determining an amount of time the projected execution exceeds the actual execution time. In one instance, the exceeding amount of time is compared against a performance threshold assigned to the VM issuing the request to determine whether to delay and/or enqueue the request. Once the subject request is released from the queue, it is conveyed to the HDD for fulfillment of the subject request.
  • the method 500 involves employing a disk-drive model to compute the projected execution time for the HDD to carry out a first request.
  • the method 500 may then involve measuring the actual execution time taken by the HDD to carry out the first request, as indicated at block 512 .
  • the disk-drive model considers a time for a disk head (used to read data from or write data to a hard disk of the HDD) to travel from a first sector of data targeted by the first request and a second sector of data targeted by a second request.
  • the projected execution time is compared against the actual execution time and the rate of the requests allowed to pass to the HDD is managed as a function of an outcome of the comparison.
  • the process of managing includes imposing a delay upon the second request before passing the second request to the HDD when the projected execution time exceeds the actual execution time by a value greater than a performance threshold.
  • the process of managing includes passing the second request to the HDD upon receipt of the second request when projected execution time does not exceed the actual execution time by a value greater than the performance threshold.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Human Computer Interaction (AREA)
  • Computer Hardware Design (AREA)
  • Quality & Reliability (AREA)
  • Debugging And Monitoring (AREA)
US13/298,993 2011-11-17 2011-11-17 Throttle disk I/O using disk drive simulation model Active 2033-04-03 US9395920B2 (en)

Priority Applications (5)

Application Number Priority Date Filing Date Title
US13/298,993 US9395920B2 (en) 2011-11-17 2011-11-17 Throttle disk I/O using disk drive simulation model
PCT/US2012/064587 WO2013074426A1 (en) 2011-11-17 2012-11-12 Throttle disk i/o using disk drive simulation model
ES12850595.5T ES2628611T3 (es) 2011-11-17 2012-11-12 E / S del disco de restricción usando el modelo de simulación de la unidad de disco
EP12850595.5A EP2780799B1 (en) 2011-11-17 2012-11-12 Throttle disk i/o using disk drive simulation model
CN201210465126.7A CN103116473B (zh) 2011-11-17 2012-11-16 使用盘驱动器模拟模型来对盘i/o扼流

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US13/298,993 US9395920B2 (en) 2011-11-17 2011-11-17 Throttle disk I/O using disk drive simulation model

Publications (2)

Publication Number Publication Date
US20130132057A1 US20130132057A1 (en) 2013-05-23
US9395920B2 true US9395920B2 (en) 2016-07-19

Family

ID=48414865

Family Applications (1)

Application Number Title Priority Date Filing Date
US13/298,993 Active 2033-04-03 US9395920B2 (en) 2011-11-17 2011-11-17 Throttle disk I/O using disk drive simulation model

Country Status (5)

Country Link
US (1) US9395920B2 (zh)
EP (1) EP2780799B1 (zh)
CN (1) CN103116473B (zh)
ES (1) ES2628611T3 (zh)
WO (1) WO2013074426A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190026142A1 (en) * 2017-07-20 2019-01-24 Fujitsu Limited Information processing apparatus, control method and storage medium
US20220229761A1 (en) * 2021-01-15 2022-07-21 Netflix, Inc. Systems and methods for optimizing hard drive throughput

Families Citing this family (38)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8346935B2 (en) 2010-01-15 2013-01-01 Joyent, Inc. Managing hardware resources by sending messages amongst servers in a data center
US8555276B2 (en) 2011-03-11 2013-10-08 Joyent, Inc. Systems and methods for transparently optimizing workloads
US8914497B1 (en) * 2011-12-15 2014-12-16 Amazon Technologies, Inc. System and method for throttling service requests having non-uniform workloads
US8782224B2 (en) * 2011-12-29 2014-07-15 Joyent, Inc. Systems and methods for time-based dynamic allocation of resource management
US8547379B2 (en) 2011-12-29 2013-10-01 Joyent, Inc. Systems, methods, and media for generating multidimensional heat maps
US8943284B2 (en) 2013-03-14 2015-01-27 Joyent, Inc. Systems and methods for integrating compute resources in a storage area network
US8881279B2 (en) 2013-03-14 2014-11-04 Joyent, Inc. Systems and methods for zone-based intrusion detection
US8826279B1 (en) 2013-03-14 2014-09-02 Joyent, Inc. Instruction set architecture for compute-based object stores
US8677359B1 (en) 2013-03-14 2014-03-18 Joyent, Inc. Compute-centric object stores and methods of use
US9104456B2 (en) 2013-03-14 2015-08-11 Joyent, Inc. Zone management of compute-centric object stores
US9092238B2 (en) 2013-03-15 2015-07-28 Joyent, Inc. Versioning schemes for compute-centric object stores
US8793688B1 (en) 2013-03-15 2014-07-29 Joyent, Inc. Systems and methods for double hulled virtualization operations
US8775485B1 (en) 2013-03-15 2014-07-08 Joyent, Inc. Object store management operations within compute-centric object stores
US9602426B2 (en) 2013-06-21 2017-03-21 Microsoft Technology Licensing, Llc Dynamic allocation of resources while considering resource reservations
US9122503B1 (en) * 2013-09-05 2015-09-01 Symantec Corporation Systems and methods for adaptive throttling of input/output requests in a virtual environment
CN104572255B (zh) * 2013-10-09 2018-09-04 中国移动通信集团公司 一种多层应用托管方法、装置及系统
CN106104414B (zh) * 2013-11-13 2019-05-21 Twc专利信托公司 存储设备以及存储和提供数据的方法
JP2015154818A (ja) * 2014-02-20 2015-08-27 任天堂株式会社 据置型のゲーム装置、ゲーム装置、ゲームシステム、コンピュータプログラム及び速度制御方法
CN104243560A (zh) * 2014-09-02 2014-12-24 赵军富 移动设备上的web服务系统
US10410155B2 (en) 2015-05-01 2019-09-10 Microsoft Technology Licensing, Llc Automatic demand-driven resource scaling for relational database-as-a-service
US10101920B2 (en) 2016-06-30 2018-10-16 Microsoft Technology Licensing, Llc Disk I/O attribution
CN110018781B (zh) * 2018-01-09 2022-06-21 阿里巴巴集团控股有限公司 磁盘流控方法、装置以及电子设备
US10552199B2 (en) * 2018-02-26 2020-02-04 Nutanix, Inc. System and method for binary throttling for live migration of virtual machines
US10552209B2 (en) * 2018-03-15 2020-02-04 Nutanix, Inc. System and method for throttling for live migration of virtual machines
US10552200B2 (en) 2018-03-22 2020-02-04 Nutanix, Inc. System and method for dynamic throttling for live migration of virtual machines
US11372580B2 (en) * 2018-08-07 2022-06-28 Marvell Asia Pte, Ltd. Enabling virtual functions on storage media
RU2718215C2 (ru) 2018-09-14 2020-03-31 Общество С Ограниченной Ответственностью "Яндекс" Система обработки данных и способ обнаружения затора в системе обработки данных
RU2731321C2 (ru) * 2018-09-14 2020-09-01 Общество С Ограниченной Ответственностью "Яндекс" Способ определения потенциальной неисправности запоминающего устройства
RU2714219C1 (ru) 2018-09-14 2020-02-13 Общество С Ограниченной Ответственностью "Яндекс" Способ и система для планирования передачи операций ввода/вывода
RU2714602C1 (ru) 2018-10-09 2020-02-18 Общество С Ограниченной Ответственностью "Яндекс" Способ и система для обработки данных
RU2721235C2 (ru) 2018-10-09 2020-05-18 Общество С Ограниченной Ответственностью "Яндекс" Способ и система для маршрутизации и выполнения транзакций
RU2711348C1 (ru) 2018-10-15 2020-01-16 Общество С Ограниченной Ответственностью "Яндекс" Способ и система для обработки запросов в распределенной базе данных
US11194620B2 (en) 2018-10-31 2021-12-07 Nutanix, Inc. Virtual machine migration task management
US11188368B2 (en) 2018-10-31 2021-11-30 Nutanix, Inc. Asynchronous workload migration control
RU2714373C1 (ru) 2018-12-13 2020-02-14 Общество С Ограниченной Ответственностью "Яндекс" Способ и система для планирования выполнения операций ввода/вывода
RU2749649C2 (ru) 2018-12-21 2021-06-16 Общество С Ограниченной Ответственностью "Яндекс" Способ и система для планирования обработки операций ввода/вывода
RU2720951C1 (ru) 2018-12-29 2020-05-15 Общество С Ограниченной Ответственностью "Яндекс" Способ и распределенная компьютерная система для обработки данных
RU2746042C1 (ru) 2019-02-06 2021-04-06 Общество С Ограниченной Ответственностью "Яндекс" Способ и система для передачи сообщения

Citations (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6490651B1 (en) * 2000-03-14 2002-12-03 Maxtor Corporation Host-based virtual disk drive for improving the performance of a hard disk drive's input/output
US20060171063A1 (en) 2005-01-28 2006-08-03 Samsung Electronics., Ltd. Seek servo control method for adapting to supply voltage variation and disk drive using the same
US20060187571A1 (en) 2004-07-29 2006-08-24 Samsung Electronics Co., Ltd. Servo controller method and apparatus for high tracks per inch hard disk drives using a delay accomodating state estimator
US20080021693A1 (en) * 2006-07-21 2008-01-24 Microsoft Corporation Storage Device Simulator
US20080168452A1 (en) * 2007-01-04 2008-07-10 Hitachi Global Storage Technologies Netherlands, B.V. Adaptive Command Scheduling
US7565484B2 (en) * 2004-06-23 2009-07-21 International Business Machines Corporation Methods, apparatus and computer programs for scheduling storage requests
US20100011182A1 (en) * 2008-07-08 2010-01-14 Hitachi Global Storage Technologies Netherlands, B. V. Techniques For Scheduling Requests For Accessing Storage Devices Using Sliding Windows
US7669044B2 (en) * 2006-09-29 2010-02-23 Microsoft Corporation Accelerated system boot
US20100070725A1 (en) * 2008-09-05 2010-03-18 Anand Prahlad Systems and methods for management of virtualization data
US7711894B1 (en) 2007-02-12 2010-05-04 Juniper Networks, Inc. Dynamic disk throttling in a wide area network optimization device
US7761875B2 (en) 2005-06-10 2010-07-20 Hewlett-Packard Development Company, L.P. Weighted proportional-share scheduler that maintains fairness in allocating shares of a resource to competing consumers when weights assigned to the consumers change
US7822920B2 (en) 2007-07-13 2010-10-26 Korea Advanced Institute Of Science And Technology Mass prefetching method for disk array
EP2333663A2 (en) 2009-12-10 2011-06-15 Empire Technology Development LLC Method and device for providing access to physical resources in a virtual machine environment
US20120278512A1 (en) * 2011-04-29 2012-11-01 International Business Machines Corporation System, Method and Program Product to Schedule Transfer of Data

Patent Citations (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6490651B1 (en) * 2000-03-14 2002-12-03 Maxtor Corporation Host-based virtual disk drive for improving the performance of a hard disk drive's input/output
US7565484B2 (en) * 2004-06-23 2009-07-21 International Business Machines Corporation Methods, apparatus and computer programs for scheduling storage requests
US20060187571A1 (en) 2004-07-29 2006-08-24 Samsung Electronics Co., Ltd. Servo controller method and apparatus for high tracks per inch hard disk drives using a delay accomodating state estimator
US20060171063A1 (en) 2005-01-28 2006-08-03 Samsung Electronics., Ltd. Seek servo control method for adapting to supply voltage variation and disk drive using the same
US7761875B2 (en) 2005-06-10 2010-07-20 Hewlett-Packard Development Company, L.P. Weighted proportional-share scheduler that maintains fairness in allocating shares of a resource to competing consumers when weights assigned to the consumers change
US20080021693A1 (en) * 2006-07-21 2008-01-24 Microsoft Corporation Storage Device Simulator
US7669044B2 (en) * 2006-09-29 2010-02-23 Microsoft Corporation Accelerated system boot
US20080168452A1 (en) * 2007-01-04 2008-07-10 Hitachi Global Storage Technologies Netherlands, B.V. Adaptive Command Scheduling
US7711894B1 (en) 2007-02-12 2010-05-04 Juniper Networks, Inc. Dynamic disk throttling in a wide area network optimization device
US7822920B2 (en) 2007-07-13 2010-10-26 Korea Advanced Institute Of Science And Technology Mass prefetching method for disk array
US20100011182A1 (en) * 2008-07-08 2010-01-14 Hitachi Global Storage Technologies Netherlands, B. V. Techniques For Scheduling Requests For Accessing Storage Devices Using Sliding Windows
US20100070725A1 (en) * 2008-09-05 2010-03-18 Anand Prahlad Systems and methods for management of virtualization data
US20140250093A1 (en) * 2008-09-05 2014-09-04 Commvault Systems, Inc. Systems and methods for management of virtualization data
EP2333663A2 (en) 2009-12-10 2011-06-15 Empire Technology Development LLC Method and device for providing access to physical resources in a virtual machine environment
US20110145814A1 (en) 2009-12-10 2011-06-16 Empire Technology Development Llc Hypervisor driver management in virtual machine environments
US20120278512A1 (en) * 2011-04-29 2012-11-01 International Business Machines Corporation System, Method and Program Product to Schedule Transfer of Data

Non-Patent Citations (23)

* Cited by examiner, † Cited by third party
Title
"European Search Report Received for European Patent Application 12850595.5", Mailed Date: May 27, 2015, 4 Pages.
"First Office Action with Search Report Issued in Chinese Patent Application No. 201210465126.7", Mailed Date: Mar. 3, 2015, 11 Pages.
A. Al Mamun et al, "Design of Controller for the Dual Stage Actuator in Hard Disk Drive Using Internal Model Approach", Proceedings of the 4th World Congress on Intelligent Control and Automation, Jun. 10 14, 2002, IEEE, pp. 2561 2565. *
A. Al Mamun et al, "Iterative Feedback Tuning (IFT) of Hard Disk Drive Head Positioning Servomechanism", The 33rd Annual Conference of the IEEE Industrial Electronics Society (IECON), Nov. 5 8, 2007, pp. 769 774. *
Ahmad, et al., "An Analysis of Disk Performance in VMware ESX Server Virtual Machines", In IEEE International Workshop on Workload Characterization, Jan. 1, 2003, pp. 65-76.
C. Ruemmler et al, "An Introduction to Disk Drive Modeling", IEEE Computer vol. 27 No. 3, Mar. 1994, pp. 17 29. *
D. Gupta et al, "DieCast: Testing Distributed Systems with an Accurate Scale Model", Proceedings of NSDI, pp. 407-421, Apr. 2008. *
DieCast: Testing Distributed Systems with an Accurate Scale Model-Published Date: Apr. 2008; D. Gupta, et al.; http://research.microsoft.com/pubs/78554/nsdi08-diecast.pdf; 15 pages.
Empirical Virtual Machine Models for Performance Guarantees; Andrew Turner, et al.; http://www.usenix.org/event/lisa10/tech/full-papers/Turner.pdf; 11 pages.
H. Fujimoto et al, "Proposal of Seeking Control of Hard Disk Drives Based on Perfect Tracking Control Using Multirate Feedforward Control", Proceedings of the 6th International Workshop on Advanced Motion Control, 2000, IEEE, pp. 74-79. *
IO Performance Prediction in Consolidated Virtualized Environments; Stephan Kraft, et al; pp. 295-306; http://www.doc.ic.ac.uk/~gcasale/icpe11io.pdf; 12 pages.
IO Performance Prediction in Consolidated Virtualized Environments; Stephan Kraft, et al; pp. 295-306; http://www.doc.ic.ac.uk/˜gcasale/icpe11io.pdf; 12 pages.
K. Fukushima et al, "Vibration Suppression PTC of Hard Disk Drives with Multirate Feedback Control", Proceedings of the 2007 American Control Conference, Jul. 11 13, 2007, IEEE, pp. 55 60. *
M. Tsuchiya et al, "Distributed Database Management Model and Validation", IEEE Transactions on Software Engineering, vol. SE 12, No. 4, Apr. 1986, pp. 511 520. *
M.S.A. Talip et al, "Knowledge Based Disk Scheduling Policy Using Fuzzy Logic", International Conference on Computer and Communication Engineering (ICCCE 2010), IEEE, May 11 13, 2010, 6 pages. *
M.Y. Javed et al, "Simulation and Performance Comparison of Four Disk Scheduling Algorithms", Proceedings of TENCON 2000, vol. 2, 2000, IEEE, pp. 10 15 vol. 2. *
Our ZFS I/O Throttle; Bill Pijewski's Blog; http://dtrace.org/blogs/wdp/2011/03/our-zfs-io-throttle; posted Mar. 1, 2011; 7 pages.
PCT International Search Report mailed Feb. 28, 2013 regarding Appln. No. PCT/US2012/064587 9 pages.
Rate Windows for Efficient Network and I/O Throttling; Kyung D. Rye, et al.; http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.24.7403&rep=rep1&type=pdf; 14 pages.
S.W. Ng, "Advances in Disk Technology: Performance Issues", IEEE Computer, vol. 31 No. 5, 1998, pp. 75 81. *
Smart Data Center: Disk IO Throttling for Optimal Performance; http://joyeurcom/2011/06/22/smartdatacenter-disk-io-throttling-for-optimal-performance; 2 pages.
T. Yamada et al, "A 16MB 3.5 Inch Ba Ferrite Flexible Disk Drive With Dual Track Following servo Modes", IEEE Transactions on Magnetics, vol. MAG 23, No. 5, Sep. 1987, pp. 2680 2682. *
Y. Gim et al, "DIG: Rapid Characterization of Modern Hard Disk Drive and It Performance Implication", Fifth IEEE International Workshop on Storage Network Architecture and Parallel I/Os, 2008, IEEE, pp. 74-83. *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190026142A1 (en) * 2017-07-20 2019-01-24 Fujitsu Limited Information processing apparatus, control method and storage medium
US10877790B2 (en) * 2017-07-20 2020-12-29 Fujitsu Limited Information processing apparatus, control method and storage medium
US20220229761A1 (en) * 2021-01-15 2022-07-21 Netflix, Inc. Systems and methods for optimizing hard drive throughput
US11899558B2 (en) * 2021-01-15 2024-02-13 Netflix, Inc. Systems and methods for optimizing hard drive throughput

Also Published As

Publication number Publication date
CN103116473B (zh) 2016-01-20
EP2780799A1 (en) 2014-09-24
WO2013074426A1 (en) 2013-05-23
EP2780799A4 (en) 2015-06-24
US20130132057A1 (en) 2013-05-23
ES2628611T3 (es) 2017-08-03
EP2780799B1 (en) 2017-04-05
CN103116473A (zh) 2013-05-22

Similar Documents

Publication Publication Date Title
US9395920B2 (en) Throttle disk I/O using disk drive simulation model
US9582311B2 (en) System, method and computer program product for data processing and system deployment in a virtual environment
US9921940B2 (en) Creating a software performance testing environment on a virtual machine system
US9648134B2 (en) Object migration between cloud environments
US9229749B2 (en) Compute and storage provisioning in a cloud environment
US8365169B1 (en) Migrating a virtual machine across processing cells connected to an interconnect that provides data communication without cache coherency support
US8621170B2 (en) System, method, and computer program product for avoiding recall operations in a tiered data storage system
US10305979B2 (en) Clone efficiency in a hybrid storage cloud environment
US20180139100A1 (en) Storage-aware dynamic placement of virtual machines
US10983913B2 (en) Securing exclusive access to a copy of a metadata track via a process while the metadata track is held in a shared mode by another process
US11074134B2 (en) Space management for snapshots of execution images
US11010295B2 (en) Asynchronous update of metadata tracks in response to a cache hit generated via an i/o operation over a bus interface
JP5987498B2 (ja) ストレージ仮想化装置、ストレージ仮想化方法及びストレージ仮想化プログラム
CN108604191B (zh) 代理对象系统
US11055017B1 (en) Throttling a point-in-time snapshot copy operation within a data consistency application
US10613986B2 (en) Adjustment of the number of tasks for a cache storage scan and destage application based on the type of elements to be destaged from the cache storage
US11188235B2 (en) Reducing data replications among storage locations
US11483205B1 (en) Defragmentation of licensed resources in a provider network
US10789008B2 (en) Reducing write collisions in data copy
US9021509B1 (en) Datacenter managed thread and multi-processing support
JP6769106B2 (ja) ストレージ制御方法、ストレージ制御プログラムおよび情報処理装置
Trivedi et al. Lessons learned: performance tuning for hadoop systems
US11983570B2 (en) Conditionally deploying a reusable group of containers for a job based on available system resources
Tran et al. Virtualizing Microsoft SQL Server 2008 R2 Using VMware vSphere 5 on Hitachi Compute Rack 220 and Hitachi Unified Storage 150 Reference Architecture Guide

Legal Events

Date Code Title Description
AS Assignment

Owner name: MICROSOFT CORPORATION, WASHINGTON

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:DENG, YIMIN;CHAU, HO YUEN;ZUO, YUE;AND OTHERS;REEL/FRAME:027246/0923

Effective date: 20111116

AS Assignment

Owner name: MICROSOFT TECHNOLOGY LICENSING, LLC, WASHINGTON

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:MICROSOFT CORPORATION;REEL/FRAME:034544/0001

Effective date: 20141014

FEPP Fee payment procedure

Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

STCF Information on status: patent grant

Free format text: PATENTED CASE

MAFP Maintenance fee payment

Free format text: PAYMENT OF MAINTENANCE FEE, 4TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1551); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

Year of fee payment: 4

MAFP Maintenance fee payment

Free format text: PAYMENT OF MAINTENANCE FEE, 8TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1552); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

Year of fee payment: 8